A computational study: reactivity difference between phosphine- and amine-catalyzed cycloadditions of allenoates and enones.

Allenoates and enones form cyclopentenes via a phosphine-catalyzed [3 + 2] cycloaddition while the amine-catalyzed [2 + 4] cycloaddition yields dihydropyrans or pyrans. The difference between these catalysts is studied with M06-2X/6-31+G* calculations. The addition of the catalyst to the allenoate is the first step in both pathways followed by the reaction with the enone. The formation of the [3 + 2] phosphorus-ylide is exergonic, and hence, the [3 + 2] cycloaddition is kinetically favored over the [2 + 4] addition. Amines do not stabilize [3 + 2] ammonium-ylides. However, electron-withdrawing groups on the enone enable [2 + 4] cycloadditions. The strength of the electron-withdrawing group further controls the α/γ regioselectivity of the [2 + 4] cycloaddition, and the analysis of the HOMO-LUMO interactions explains why only E-dihydropyrans from the direct γ-[2 + 4] cycloaddition have been observed in experiments. The quantum calculations further reveal a new path to the α-[2 + 4] product starting with an intermediate Rauhut-Currier reaction. This new path is kinetically favored over the direct amine-catalyzed α-[2 + 4] cycloaddition.